The paper provides conclusions on the impact of VAAs based on an experiment carried out in Hungary. In the framework of a panel study subjects first filled in a questionnaire, then some of them were directed to two smartvote portals. Some received advice from both, some from one, the control group visited none. In the second wave the subjective impressions of the VAA-users were collected. In the third wave, after the national parliamentary elections, the subjects were asked whether they voted and if yes then for which party. The complex experiment proved that the two VAA projects, in spite of the many differences in their structure, provided visitors with roughly similar pieces of advice. According to the subjective recollections seven per cent of the users changed their vote intentions in line with the received advice. But the third wave of the panel study failed to confirm the existence of significant number of party switchers, and there was no proof for the mobilizing or demobilizing effects either. It turned out, however, that those who received identical advice from the two websites went to vote in significantly higher numbers.
